Cloud - Bare Metal Infrastructure Engineer

ABOUT THE ROLE

Most infrastructure engineers never get to build a cloud from the metal up. At Render, you will. Developers and businesses around the world run their apps on Render, and we are building our own bare metal footprint to give them better performance and economics. You'll help take our first colocation site from an empty cage to production workloads, then scale that foundation as we expand our physical presence to thousands of hosts over the coming years.

You'll join a small, experienced team with real ownership. You'll partner closely with network engineering and our Kubernetes platform team, and the decisions you make now will become the template for every site that follows.

WHAT YOU'LL DO

- Take our first colocation site from empty cage to production workloads, then scale what you build as we grow to more sites and thousands of hosts

- Design how servers go from the loading dock to production with minimal human touch: netboot, imaging, BMC-driven discovery, and host identity

- Partner with network engineering on what the fleet needs from the network, from out-of-band management to production connectivity

- Build the interface our Kubernetes platform team uses to request capacity, see host health, and coordinate maintenance, so consuming bare metal feels as easy as consuming cloud

- Own the fleet's source of truth for hardware, software, and the physical environment

- Automate the full server lifecycle, from burn-in and acceptance testing through remediation, staged firmware rollout, and RMA

- Build automation like a production system, with guardrails on destructive actions and designs that survive site-level failure

- Shape the technical and operational direction of our bare metal platform, from build-vs-buy decisions to how we operate thousands of hosts, and help hire the team that runs it

WHAT WE'RE LOOKING FOR

- Hands-on experience building and operating automation for bare metal fleets in datacenters or colocation

- Both sides of the work: time with physical hardware and colo providers, and modern, software-driven fleet management

- Strong Go or Python, and experience designing systems with reconciliation loops and idempotent, failure-tolerant actions

- Depth below the abstraction: you know what your tools are doing under the hood and can debug them when they fail

- Experience operating through real failures, such as bad hardware batches, flapping BMCs, or a rollout you had to halt

- Deep Linux knowledge and solid networking fundamentals (DHCP, PXE, DNS, BGP)

- Working knowledge of Kubernetes, enough to design the interface between the fleet and the clusters that run on it

- Clear communication, comfort with ambiguity, and the willingness to push back on a plan when something doesn't add up

NICE-TO-HAVES

- Operating bare metal fleets at a scale of thousands of hosts

- Experience on a bare metal or datacenter build-out from its early days

- Running a hybrid environment, or moving workloads between public cloud and owned hardware

- DCIM/IPAM tools such as NetBox or Nautobot, or provisioning stacks such as Ironic/Metal3, Tinkerbell, or MAAS

- Building reconcilers or state machines for physical infrastructure

- Hardware depth: component diagnostics, burn-in tooling, firmware, or Redfish and IPMI at fleet scale

- Building internal platforms or APIs used by other infrastructure teams

- Based near our colocation sites, or able to visit them regularly